In the United States, the Food and Drug Administration (FDA) is charged with regulating all pharmaceutical marketing through its Division of Drug Marketing and Advertising (DDMAC). However, ensuring advertisement adherence is challenging and FDA resources dedicated to this task are limited. DDMAC's fiscal year 2008 budget of $9 million [13] is dwarfed by the pharmaceutical industry's $58 billion marketing budget [14]. While the FDA can require that companies change or remove non-adherent advertisements, given DDMAC's size and the volume of marketing materials it regulates, it would be unreasonable to expect each individual pharmaceutical advertisement to be reviewed.

Acknowledging this regulatory challenge, the FDA has recently asked physicians to report non-adherent or misleading advertisements through its “Bad Ad” program, explaining that “With your valuable assistance, FDA can be more effective in limiting the number of misleading promotional messages directed to health care professionals” [15]. This program may prove useful in identifying grossly misleading advertisements; however there has been no systematic assessment of the adherence of physician-directed advertisements to FDA regulations in the last 20 years [16] and none to current standards, which were last updated in 2001.

Despite the high rates of FDA non-adherence, the mean number of biased features in each advertisement was low and most advertisements we reviewed satisfied the majority of FDA-guidelines despite a general absence of content important for safe prescribing. FDA guidelines, though detailed, do not target many of the ad features most important for providing prescribers with useful information. For instance, FDA emphasizes avoiding frankly false information and balancing efficacy and safety information [19] but does little to encourage the presentation of useful and accurate information. An advertisement containing no specific efficacy claim, no quantification of drug safety and no verifiable references would adhere fully to FDA guidelines, despite presenting no practical information for clinicians. We found many such advertisements in our sample; for example an advertisement for a new chemotherapeutic agent presented the indication, the claim that it was the first in a new drug class, a vague image, and safety information with no quantification of effect and no references.

The expectation that advertisements should inform rational prescribing by presenting complete drug safety and efficacy information may be unrealistic since advertisements primarily serve a marketing function [20] and are not designed to train physicians to properly prescribe. However, proponents of physician-directed advertising emphasize the important role advertisements play in educating physicians and other prescribers about new drugs [3], and it is clear that advertisements do serve to influence prescribing by providing information to physicians. While the majority of physicians deny that advertisements inform their prescribing [21], marketing research has consistently shown that journal advertising is the most profitable form of drug marketing, with an estimated return on investment of $5 for every dollar spent [1]. Furthermore, in some cases exposure to physician-directed advertising has been shown to be associated with less effective, lower quality prescribing decisions [22]. It is clear that prescribers are ultimately responsible for the safety and efficacy of drugs they prescribe but few doctors recognize the extent to which they are influenced [23]. Physicians should ensure that their prescribing is informed by the clinical literature and not by marketing materials. However, given the high risk associated with many advertised drugs, any tendency of advertisements to encourage inappropriate prescribing may pose dangers to patients. If FDA is truly committed to improving prescribing and protecting the health of the public it should demand that content important for safe and effective prescribing be consistently presented in physician-directed advertisements.

A major challenge of our study was performing advertisement evaluations given the subjectivity of the current FDA guidelines. For instance, the FDA describes that advertisements are misleading if they use “a quote or paraphrase out of context to convey a false or misleading idea” or “headline, subheadline, or pictoral or other graphic matter in a way that is misleading”, without providing explicit definitions. We created objective criteria for many items but scoring still required subjective determinations. Our conservative approach to subjective items may have underestimated the true non-adherence rate of the advertisements we evaluated, which may explain why the rate of advertisement non-adherence in our study is lower than the 92% non-adherence rate demonstrated in a 1990 study of journal advertisements directed at physicians [15].

There are several other important limitations to our study. First, we sampled advertisements from a single month in 2008, so we cannot comment on trends in advertisement characteristics over time. We focused on high-impact biomedical journals publishing original research, excluding lower-impact and more narrowly focused publications and those publishing clinical reviews, which may also be frequently read among clinicians. We may have missed important differences in advertisements published in these other types of publications and we expect the volume of advertisements to have been greater in these journals. However, we suspect that major differences in advertising content are unlikely since pharmaceutical company advertisements are generally created for use in many journals, not just the highest-impact journals. We did not evaluate advertisements presented in other contexts (e.g. direct mailings to physicians) or other media (e.g. online). Data on the influence of newer marketing strategies is limited but journal advertisements appear to remain influential [24]. In addition, while visual aspects of advertisements may be highly influential, a full content analysis of advertisement imagery was beyond the scope of this paper. We did note a few patient photos which were misleading enough to lead to FDA non-adherence, but it is likely that other advertisements utilized more subtle imagery to imply more efficacy or broader patient applicability; thus our analysis may underestimate the misleading nature of some advertisements. Further, the items we measured to evaluate information important for safe prescribing were not validated, although the importance of quantifying benefits and risks is generally accepted [18] and the concepts evaluating safe prescribing are straightforward.

Our findings have important policy implications. The FDA has demonstrated a desire to improve the quality of pharmaceutical advertisements, most recently by enlisting doctors to review advertisements through its “Bad Ad” program. FDA should now move to update and simplify its guidelines for physician advertisements to facilitate the review process. Updated FDA guidelines should be straightforward and objective, and should include requirements that physician advertisements present clear risk quantification, absolute benefit information, description of the appropriate population to receive the drug, and verifiable references, either to the peer-reviewed published literature or to studies registered within federally-sponsored internet sites such as ClinicalTrials.gov in the case of unpublished data. These new guidelines might be applicable only to physician-directed advertisements, with separate guidance for DTC material, with greater focus on safety issues. Enforcement would still be necessary, as in other nations advertisements have failed to adhere to new straightforward requirements [25]. However, new objective requirements would serve to enhance prescribing and public health by improving the quality, accuracy and transparency of advertisements, simplify at least part of the FDA review process, and facilitate physician participation in reporting “bad” advertisements.

Table S1

Figure S1